Quantitative Analysis of Coon's Taxonomy

Carleton S. Coon's Caucasoid subracial taxonomy is an example of a typological system. Of course we can't be sure how accurate his system is unless we see data for the populations it describes. But we can also look at the type examples chosen by Coon and ask whether or not they segregate into the groups defined by Coon, thus checking the internal consistency of his taxonomy.(*)

In order to answer this question I used 152 of Coon's individuals from 40 of his plates. I averaged the metrical data on each of these plates and thus created "average types". Subsequently the data was standardized by subtracting the mean and dividing with the standard deviation and C-scores were extracted by subtracting the mean of each row (individual) from the measurements. Using this, I constructed a dendrogram based on Euclidean distance:

Before we continue, let's summarize Coon's system. Coon thinks the main division in Caucasoids is between narrow-faced Mediterraneans and broad-faced Upper Paleolithics. The former are rather gracile, while the latter are robust. Mediterraneans are divided into various groups based on their size and pigmentation and other traits, but are presumed to form a single racial family. Upper Paleolithics are also divided into subgroups, most notably gracilized and brachycephalized Alpines and massive dolichocephalic Bruenns and brachycephalic Borrebies. Additionally there are races formed by the admixture of Mediterraneans with Alpines or Borrebies: the Dinarics and Armenoids; these have long faces beaky noses and short heads (think Count Dracula).

I've annotated the dendrogram above with numbers to make the discussion easier. The examples segregate into two groups. 1 is mainly broad-faced and 2 is mainly narrow-faced. So, Coon's taxonomy is self-consistant so far. There are of course some exceptions. Group 1 separates into 3, which includes mostly UP types and 4 which includes mostly Alpines and Alpine-derivatives. Again, Coon's system seems to be fairly consistent in dividing the broad-faced people into unreduced and reduced varieties.

The narrow-faced cluster 1 is divided into brachycephalics 5 which correspond to Dinarics and dolichocephalics 6 which correspond to Nordics and Mediterraneans. Coon actually took on the view after publishing the Races of Europe that "Dinarics", which he formerly thought were the result of admixture with Alpines, were actually the result of cradling practices. No matter what the origin of the Dinarics, they seem to be in most respects more similar to the Mediterraneans.

Let's go into one level deeper. Cluster 4, which we labeled "Alpine" is divided into 7 and 8. 7 includes "Danubians", aberrant Alpine forms, Pontic Mediterraneans, "altered" Nordics and others. They probably correspond to "Mixed Alpine" forms that combine Mediterranean-Alpine forms. Cluster 8 includes "Borreby" forms, Neo-Danubians, East Baltics and Alpines. We are tempted to think that the distinction between "Borrebies" and Alpines is from a metric perspective artificial.

The subcluster of 6, the "Nordic-Mediterraneans" are 9 which comprises of most Europeans, and 10 which comprises of most Non-Europeans and (surprisingly) Scandinavian long-headed Bruenns. Bruenns are distinguished mainly by their robustness and big size. But are they a valid group? This analysis seems to indicate that they are not, in agreement perhaps with the opinion of Earnest Hooton who considered the UP dolichocephals as early Mediterraneans. This result is somewhat surprising since the UP dolichocephals, who are thought to be the most ancient stratum of the European population cluster with the Asiatic Mediterraneans. Metric variation of course can't tell the whole story, but at most we can say that the Bruenns are not distinctive enough metrically or distinguished by a "wide racial gap" from Mediterraneans as Coon thought.

In summary: Coon's typological system was analyzed from a metric perspective. The main distinction is not between Europeans and non-Europeans, or Upper Paleolithics and Mediterraneans, it is rather between broad-faced and narrow-faced forms. The chiefly Mediterranean character of the Dinarics is confirmed, as is the metric unity of Mediterraneans and Nordics. The distinction between reduced and unreduced varieties gave contradictory results and does not in my opinion merit the attention that Coon gave it in his system.

(*) With the caveat that Coon's types are defined on the basis of metrical, morphological and pigmentation data. Here I only deal with the metrical aspect.

Posted by Dienekes at December 14, 2003 11:30 PM | PermaLink